3D Data Denoising via Nonlocal Means Filter by Using Parallel GPU Strategies

Table 9

Full unrolling algorithms for different video formats. Observe that the execution times include the time needed to copy data from CPU to GPU and from GPU to CPU.

Video size	MB	Execution time
		(11^3,3^3)	(21^3,3^3)	(11^3,5^3)	(21^3,5^3)
PAL ×50	79.1	16.8	85.8	55	276
PAL ×100	158	38.8	239	131	807
PAL ×200	316	82.9	545	284	1868
PAL ×500	791	215	1463	742	5052
HD ×50	148	31.4	161	103	517
HD ×100	297	72.7	448	246	1523
HD ×200	593	155	1022	532	3502
HD ×500	1483	403	2743	1391	9472
FULL-HD ×50	396	78.5	402	257	1294
FULL-HD ×100	791	182	1119	615	3781
FULL-HD ×200	1582	388	2550	1327	8738
FULL-HD ×500	3955	1007	6846	3473	23634